透過您的圖書館登入
IP:3.88.60.5
  • 學位論文

電鍍銅銦硒薄膜之材料特性、成長行為控制及電化學結晶行為-應用於銅銦硒太陽能電池

Materials characteristics, growth control, electrocrystallization behaviors of the electrodeposited CuInSe2 for solar cell applications

指導教授 : 林招松
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


本論文中利用單槽電鍍法於銅銦硒水溶液系統中製備出具有平整、緻密且符合劑量比之銅銦硒太陽能電池吸收層材料,而製程中的陰極電位、氯化鉀濃度(0 ~ 500 mmol L-1)、濺鍍鉬基板、鉬基板前處理製程均被證實對所製備的銅銦硒薄膜性質與後硒化結果有重要的影響。另外,不同電位下(–0.65 ~ –0.80 VSCE)對銅銦硒化合物之電化學結晶行為與相對應之組成變化也被詳細而有系統地探討。實驗中並嘗試以單槽電鍍銅銦硒薄膜為基礎製作銅銦硒薄膜型太陽能電池,以求能透過電鍍製程的改善進一步提升電池的發電效率及製程穩定性。 實驗結果發現在適當濃度的電鍍液中,ex: CuCl2, InCl3, H2SeO3, KCl 進行定電位還原,並將還原電位控制在–0.5 ~ –0.8 VSCE 左右時,鍍層含量可控制在Cu/In= 0.8 ~ 1.0 及Se/Cu+In+Se ≈ 0.5 左右,含量控制雖然在製作太陽能電池吸收層之P-N junction 中扮演著電性調控的重要角色,但是高平整性、高鍍層緻密性與成分均勻性才是電鍍法製作銅銦硒薄膜應用於高品質太陽能電池中最重要的條件,而這樣的材料特性在本研究中可藉由選擇適當的還原電位、控制鍍液氯化鉀濃度(0 ~ 300 mmol L-1)、或利用Mo 基板的改質製程來達到。氯化鉀一般認為是作為鍍液的導電鹽使用,並不會參予參與陰極反應,然而隨著氯化鉀濃度的上升,溶液中的活性物種ex: CuCl2, InCl3, H2SeO3 於pH1.55 水溶液中的穩定態,可透過不同機構的引導下,其還原量會產生差異,並導致銅銦硒薄膜成份偏離、在局部區域出現Cu2-xSe 二次相的比例上升、鍍層形貌從平整生長轉向結瘤狀或樹枝狀生長之結果。雖然在鍍層品質會因氯化鉀濃度提高而降低,但隨著氯化鉀濃度高於300 mmol L-1,Cl–與In3+會錯合成InCl3-nn ,降低銦硒氫氧化物沉澱機率,並提升鍍浴穩定性。另外,一般電鍍薄膜甚少會隨底材的變化而改變,然而研究發現濺鍍Mo 基板對電鍍銅銦硒薄膜會帶來明顯且重要的改變,尤其是在為了維持鍍浴穩定的情況下,仍可透過控制基板條件保持高品質的銅銦硒薄膜,結果顯示濺鍍條件除了影響Mo 基板結晶性外,將此基板做為工作電極置於銅銦硒電鍍液中,其開路電位亦會有系列性的改變,因此在定電位還原的情況下,不同Mo基板之過電位的差異進而影響銅銦硒薄膜的成分。此外不同濺鍍Mo 基板含氧量的變化,亦會大大影響電鍍時的析氫效率,高含氧量(29 at %)的Mo 基版將有利於生成平整的銅銦硒薄膜。然而由於結晶性較差或是高含氧量的Mo 基板電阻值較高,ex: 電阻值控制在13 ~ 580 μΩ cm,因此嘗試利用Mo 基板的表面改質或電鍍條件的調控,達成平整的電鍍CIS 鍍層,並維持低電阻之背電極特性,其製程選擇包括了:利用真空製程製作MoOx/Mo bilayer、300 oC 之熱空氣氧化、及酸性HCl 溶液鈍化、多段式電位電鍍法於新鮮Mo 表面。而實驗結果也證實確實有應用上的效益。 以電鍍法為基礎成長的銅銦硒結晶行為目前是較為開放的一個課題,本論文首次利用TEM 碳膜上進行電鍍實驗,並利用電子顯微鏡直接分析成長初期化合物結晶成長行為,另外並透過數值分析解析其計時電流曲線,發現其為2D 瞬時成核與3D 瞬時成核的疊加,從High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM)及EDS 成分分析結果顯示銅銦硒在電鍍發生的初期4 ms,Cu2-xSe 首先生成,並無任何元素態的Se、Cu、In 等發現,經過48 ms 後,Cu2-xSe 表面的Cu2+及H2SeO3 濃度下降,相對In3+離子濃度增加,最後在Cu2-xSe 核點外側,誘發CuInSe2 還原並生成核殼結構,開始主導整個電鍍過程。

並列摘要


The compact, smooth, with an expectative composition CuInSe2 electrodeposits used in CIS solar cell were successfully prepared by single-bath CIS solution. Reduction potential, KCl concentration (0 ~ 500 mmol L-1), sputtering Mo coated glass, and pretreatment of Mo substrate were found to significantly influence the properties of the deposited CIS and selenized CIS. In addition, the electrocrystallization of CIS at different reduction potentials (–0.65 ~ –0.80 VSCE) and corresponding the composition were investigated. This study also attempted to fabricate a CIS solar cell based on the single-bath electrodeposition process for CIS adsorption layer. The improvements of the cell efficiency and reproducibility of electrodeposited CIS layer due to the modified processes were explored as well. The experimental results indicated that the ratio of Cu to In between 0.8 ~ 1.0 and the ratio of Se to CIS slightly above 0.5 can be obtained potentiostatically at –0.5 ~ –0.8 VSCE in the selected CIS solution composed of CuCl2, InCl3, H2SeO3 and KCl. CIS composition plays an important role in the electric properties of PN junction for the solar cell application. However, smooth surface with compact structure and uniform composition were confirmed to be the most important factors that led to the high quality adsorption later for the fabrication of CIS solar cell, especially for the electrodeposition process of CIS. Furthermore, these kinds of materials characteristics of electrodeposited CIS were preferably obtained at optimal deposition potential, appropriate KCl concentration (0 ~ 300 mmol L-1), “passivative” Mo substrate and multi-step potential deposition. In general, KCl has been widely used as a conductivity salt owing to its great ionic mobility and does not participate in the cathodic reaction. However, in CIS deposition, KCl was found to enhance the reduction current of individual electroactive species (ex: predominated ion of added salts, likes CuCl2, InCl3, H2SeO3 at pH1.55) in varying degrees by different mechanism. Thus, the electrodeposited CIS results in the composition deviation from stoichiometry, appearance of secondary Cu2-xSe phases in local areas causing degradation of the device performance, and compact structure changed to granular or dendritic growth of CIS. Accordingly, the quality of CIS deposits were deteriorated by increasing KCl, though the stability of CIS solution was promoted as the KCl concentration was above 300 mmol L-1 due to the competitive complexing; that is, complexes are formed between indium ions and chloride ion (InCl3-nn ) to avoid the formation of indium selenide hydroxide suspensions. Furthermore, in common metal electrodeposition, the deposits are hardly influenced by different substrates. However, the sputtering Mo coatings strongly influenced the composition and morphology of the CIS deposits, especially, the deposits obtained from the higher KCl concentration for the purpose of stable electrolyte and without loosing the smooth, compact CIS sturcuture. Experimental results indicated that the sputter parameters influence the crystallinity, and the crystallinity of Mo coatings correlated very well of the open circuit potential in the electrolyte for CIS electrodeposition. Consequently, the Cu/In ratio of CIS deposits plated at a constant potential of –0.7 VSCE varied with the distinct Mo coatings of different overpotentals. Various degree of hydrogen evolution reaction (HER) can also be retarded by the different amoumt of oxygen content in the Mo coatings. Higher content of oxygen of 29 at % can resulted in a compact CIS deposit. However the Mo coatings deficient in crystallinities or with higher oxygens content show higher conductivity were generally unfavorable to the electron transfer. Ex: the condictiviy of distinct Mo substrate can vary within 13 ~ 580 μΩ cm. Therefore, the surface modifications of high conductivity Mo substrates and the newly multi-step deposition were investigated to achieve the compact CIS deposits without sacrificing the electric conductivity. These processes included: MoOx/Mo bilayer manufactured in vacuum process, oxidation in air atmosphere at 300 oC, Mo passivated in HCl solution, potentiostatic deposition with multi-step. The experimental results have also been established. However, the study of the electrocrystallization process in CIS electrodeposition is still absent. Here, we employed a new and straightforward approach to study the early electrocrystallization using carbon-coated nickel TEM grids as the substrate for CuInSe2 electroplating and directly observed it in the TEM. The potentiostatic current transients of the peak associated with CuInSe2 were also evaluated by numerical analysis and showed that the initial stage of deposition included a 2D instantaneous nucleation overlapped with a 3D instantaneous nucleation. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and EDS reveals that nanocrystalline cubic structure Cu2-xSe is deposited within the first 4 ms. No elementary Cu, In, Se can be detected. After 48 ms, as Cu2+ and H2SeO3 around Cu2-xSe nuclei become depleted, In+3 is then reduced and the tetragonal CuInSe2 nanocrystals start to deposit around the Cu2-xSe nuclei surface to form core-shell structures.

參考文獻


[29] S.O. Kasap, Optoelectronics and Photonics–Principles and Practices, Prentice
[16] A.M. Gabor, J.R. Tuttle, D.S. Albin, M.A. Contreras, R. Noufi, Appl. Phys.
[23] R.N. Bhattacharya, J.F. Hiltner, W. Batchelor, M.A. Contreras, R.N. Noufi, J.R.
[82] K. Ramanathan, F.S. Hasoon, S. Smith, D.L. Young, M.A. Contreras, P.K.
[117] R.C. Valderrama, M. Miranda-Hernandez, P.J. Sebastian, A.L. Ocampo,

延伸閱讀